Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

In this paper, we present a novel image retrieval technique based on salient regions that are invariant under viewpoint and illumination variations. The salient regions are detected according to local entropy and scale selection. The detected regions have very high repeatability under various viewpoint and illumination changes. We apply the invariant region detector on content-based image retrieval applications. The detected regions are the most informative ones, hence are potentially more effective for image indexing and retrieval. Generalized colour moment invariants are used as the invariant descriptor for characterizing the selected salient regions under geometric and photometric changes. Two datasets are used for experiments: one for evaluating the invariance of the algorithm under geometric and photometric transformations; the other for testing the algorithm on object category retrieval. The experimental results show that our proposed region detector is very efficient and effective on retrieving a variety of cluttered images with partial occlusion. © 2006 Elsevier Inc. All rights reserved.

Original publication




Journal article


Journal of Visual Communication and Image Representation

Publication Date





1256 - 1272